ABSTRACT Families with children are the fastest growing subgroup of homeless households in the United States, now comprising over one-third of the homeless population.1,2 Homelessness and associated adversities faced by these extremely poor families threaten child behavioral health, creating complex needs in the context of scarce resources.3?5 Estimates suggest one-third to over one-half of homeless children experience clinically significant internalizing or externalizing behavior problems despite high intensity of homeless services receipt3,4,6 Limited empirical research has investigated how homeless families prioritize child behavioral health, or how shelter providers facilitate caregivers in supporting child behavioral health.4,7?9 Failure to align caregiver priorities with provider policies can have system-level impacts by leaving needs unmet, contributing to ongoing service needs, and increasing strain on providers over time. Understanding and intervening on the complexity of needs among homeless families requires conceptual approaches that engage multiple stakeholders and systems thinking, incorporating feedback mechanisms and adaptation over time to support child behavioral health, promote return to stable housing, and reduce shelter reentry.10?13 Using both qualitative data from stakeholders and administrative homeless services data, the present mixed methods study applies a systems perspective10,11 in three phases to understand the decision-making processes that drive service utilization patterns of homeless families with children. First, group model building sessions?structured focus groups that apply systems thinking to develop qualitative theoretical models14,15?will be conducted with homeless caregivers of children and homeless service providers; these sessions will increase understanding of how caregivers prioritize child behavioral health needs in the context of homelessness, how providers facilitate caregivers to support child behavioral health, and the relationship to homeless service use patterns. Second, the knowledge generated from group model building will guide the building of a system dynamics simulation model using homeless services administrative data to track current patterns of family homeless services use while incorporating caregiver priorities and provider decision-making processes. Finally, a second set of group model building sessions will generate potential interventions; these will be built into the simulation model at potential leverage points to test whether homeless service system outcomes can be improved over time. The innovative application of system dynamics to family homeless service delivery in the proposed study will inform efforts to promote child behavioral health among vulnerable families. Through the proposed training plan, the PI will develop valuable expertise in studying homeless services and child behavioral health, applying community-based system dynamics methodologies, and developing academic-community research partnerships necessary to pursue a research career as a tenure-track professor.